Single-cell and Bulk RNA-seq Analysis

We used single-cell and bulk RNA-seq data generated by Kim et al. [28 (link)], which have been deposited in the Sequence Read Archive (SRA, NCBI) using the Gene Expression Omnibus (GEO) (GSE69405). The RNA-seq data for the primary matched normal (pN) and mouse xenograft (xenoT) samples can be accessed in the GEO by GSE70968. According to Kim et al. [28 (link)], single-cell capture and cDNA amplification were performed using a SMARTer kit (Clontech, Mountain View, CA) on a 17–25 μm microfluidic chip (Fluidigm, San Francisco, CA). Sequencing libraries were generated using a Nextera XT DNA Sample Prep Kit (Illumina, San Diego, CA) for single cells or a TruSeq RNA Sample Preparation v2 kit (Illumina) for bulk RNA samples, and sequencing reads were generated using HiSeq2500 sequencing system in the 100-bp paired-end mode using a TruSeq Rapid PE Cluster kit and a TruSeq Rapid SBS kit (Illumina). RNA-seq reads were filtered at Q33 using Trimmomatic-0.30 [30 (link)] and were aligned to a human genome reference (hg19) using Tophat2 [31 (link)]. Transcripts were estimated as fragments per kilobase of transcript per million mapped reads (FPKM) using Cufflinks [31 (link)].